Novel technique for simultaneous imaging of the breast stiffness and incompressibility using quasi-static elastography
Keywords:Ultrasound, Breast Cancer, Imaging, Machine Learning, Image Processing
Breast cancer is one of the most common cancers,
representing 25% of all new cancers and 13% of all cancer
related deaths in Canadian women . Early detection before
treatment of breast cancer is paramount as survival rates decrease significantly over time. Some of the most common diagnostic and screening procedures include breast MRI, mammography, and manual examination. These methods are
either too costly or have difficulty detecting or differentiating
malignant tumors from benign ones without a follow-up biopsy. One technique that has shown a potential to minimize
the number of biopsy cases is ultrasound elastography (USE),
which images the breast stiffness which is known to be substantially different for normal and pathological tissue .
Currently, the images produced by USE tend to be
of low quality, plagued by noise and distortions due to the
nature of ultrasound, inconsistent mechanical stimulation by
the operator, and other inconsistencies in acquisition or tissue
structure. We have developed new real-time techniques
aimed at improving the data quality by enforcing known
physical properties . This work aims at developing a
method by which the Young’s modulus, shear modulus, and
Poisson’s ratio are simultaneously reconstructed.
II. PROPOSED METHODS
The reconstruction algorithms follow an iterative
technique that include stress calculation followed by
Young’s modulus and shear modulus reconstruction based on
Hooke’s law where both axial and lateral strains are utilized.
The Poisson’s ratio is then constructed using the two moduli
images. By running these algorithms on a dataset of several
in-silico models of tissue deformation at different loading
levels, along with clinical breast cancer cases, we analyzed
the accuracy, signal to noise ratio, and contrast to noise ratio
of the stiffness images to determine which algorithms are
more suited to breast cancer diagnosis.
The results of both in-silico and clinical cases show that,
measured by their high signal-to-noise ratio and contrast-tonoise ratio, the reconstructed images of Young’s modulus,
shear modulus and Poisson’s ratio are of high quality. Each
image shows some unique features. For example, while the
Young’s modulus measures tissue stiffness the shear modulus can be used to assess bonding between adjacent types of
tissue. The latter shows low shear modulus at the outline of
The investigation shows promising capability of the
proposed algorithms to produce high quality images of tissue
Young’s modulus, shear modulus and Poisson’s ratio that
showed complementary features that can be fused for accurate breast cancer diagnosis. Further investigation is necessary to measure the method’s sensitivity and specificity.